We ask whether conformity, copying the most observed behavior in a population, can affect free riding in a public goods situation. Our model suggests that, if free riding is sufficiently frequent at the start of a public goods game, conformity will increase the growth rate of free riding. We confirm this prediction in the experimental lab by showing that more free riding occurs when players have information about the distribution of contributions than when players know only the aggregate contribution level. As a stricter test, we econometrically estimate the dynamic on which the model is based and find that, controlling for the payoff incentive to free ride, players react significantly to the number of free riders in their groups. Further, conformity is significantly stronger when players have more information about the choices of others.